Tmb Standardization By Alignment To Reference Standards: Phase Ii Of The Friends Of Cancer Research Tmb Harmonization Project.

2624 Background: Tumor mutational burden (TMB) is a predictive biomarker of response to immune checkpoint inhibitors across multiple cancers. In Phase 1 of the Friends of Cancer Research Harmonization Project, we demonstrated a robust correlation between TMB estimated using targeted next-generation sequencing (NGS) gene panels and whole exome sequencing (WES) applied to MC3-TCGA data. These findings demonstrated variability in TMB estimates across different panels. Phase 2 evaluates sustainable TMB reference standard materials for TMB alignment to assess this variability. The goal of this effort is to establish best practices for estimating TMB in order to improve consistency across panels, for the sake of optimizing clinical application and facilitating integration of datasets generated from multiple assays. Methods: Fifteen laboratories with targeted panels at different stages of development participated. We identified a set of reference standards consisting of 10 well-characterized human-derived lung and breast tumor-normal matched cell lines. WES was performed using a uniform bioinformatics pipeline agreed upon by all team members (WES-TMB). Each laboratory used their own sequencing and bioinformatics pipelines (tumor-only and tumor-normal) to estimate TMB according to genes represented in their respective panels (panel-TMB). The association between WES-TMB and each panel-TMB was investigated using regression analyses. Bias (relative to WES-TMB) and variability in TMB estimates across panels were rigorously assessed. All analyses were blinded. Results: The set of reference standards spanned a clinically meaningful TMB range (4.3 to 31.4 mut/Mb). Preliminary data from 12 laboratories shows a good correlation between panel-TMB and WES-TMB in this empirical analysis. Across panels, regression R 2 values range 0.77-0.96 with slopes ranging 0.60-1.26. Calibration analyses that seek to minimize variability of TMB estimates across panels using the established set of reference standards are ongoing, as well as investigating cancer type dependence on the relationship between panel-TMB vs. WES-TMB, which will be available at the time of presentation. Conclusions: Preliminary findings demonstrate feasibility of using sustainable reference control cell lines to standardize and align estimation of TMB across different targeted NGS assays. Future studies aim to validate reference standard material as a reliable alignment tool by using formalin-fixed paraffin-embedded human tumor samples.